[OSM-legal-talk] sharealike trigger

Paul Norman penorman at mac.com
Mon Jul 22 10:46:17 UTC 2013

> From: John Bazik [mailto:jbazik at gmail.com]
> Sent: Sunday, July 21, 2013 10:19 PM
> Subject: Re: [OSM-legal-talk] sharealike trigger
> What consitutes substantial?  I've read many threads on this, but I 
> find myself no more able to determine what that might be.

If there's ambiguity about the term substantial, it's coming from the
relevant laws. The ODbL is just echoing the database directive.
Insubstantial is much like fair use. The license should not attempt to
define these terms.

I don't see substantial as relevant for most use-cases discussed on the
list, they have been substantial in quantity.

> And the sharealike trigger is pulled whether the substantial data is 
> already in OSM or isn't, but "could be."  Technically, any data that 
> references an OSM foreign key could be in OSM.

It sounds like you're confusing computer science and RDMS databases with the
legal concept of a database. For the rest of this message, I'm talking about
the legal concept of a database.

> Routes, "in some cases," are of interest.  What if the only meta 
> information associated with routes is the users of another service and 
> their opinions?  Giving up user account information is obviously 
> problematic for any organization, commercial or non-profit.

Well there's a pretty strong precedent by the largest user of OSM data to
not consider user data part of the same database as the map data: osm.org
itself. osm.org users are more closely tied to the map database then any
external service is likely to be, but they're treated as a collective

If they were one database then the user info would have to be distributed
under the ODbL, or if the user data couldn't be distributed due to other
laws (e.g. privacy laws), then the map data couldn't be distributed at all.

Off hand, I can think of several databases stored in the same RDMS
on osm.org

- user database
- notes database
- map database
- GPX database
- message database
- oauth client database

> I fear the definition of a derivative work is akin to Justice Potter's 
> famous construction, "I know it when I see it."

The real question is when is something one database and when is it multiple
collective databases. The gray area is a lot smaller than some people

More information about the legal-talk mailing list